Skip to content

CI: Add scheduled release workflow - #371

Merged
cgwalters merged 1 commit into
composefs:mainfrom
jeckersb:scheduled-release
Aug 6, 2026
Merged

CI: Add scheduled release workflow#371
cgwalters merged 1 commit into
composefs:mainfrom
jeckersb:scheduled-release

Conversation

@jeckersb

@jeckersb jeckersb commented Aug 5, 2026

Copy link
Copy Markdown
Collaborator

Adapted from the bootc-dev/bootc scheduled release workflow.

Signed-off-by: John Eckersberg jeckersb@redhat.com

Adapted from the bootc-dev/bootc scheduled release workflow.

Signed-off-by: John Eckersberg <jeckersb@redhat.com>
@jeckersb

jeckersb commented Aug 5, 2026

Copy link
Copy Markdown
Collaborator Author

Notable differences between this and the bootc version, didn't want to put this in the commit message incase we want to change any of it:

  • bootc runs every Monday, this runs on the 1st of the month, weekly feels excessive here (can always run manually if desired)
  • I removed all of the app token stuff which doesn't seem to be necessary anymore? Per https://github.com/peter-evans/create-pull-request#token I think we can just use the default GITHUB_TOKEN. Not at all positive about that though.
  • I specifically excluded composefs-capi from the version bump since it doesn't match the rest of it and I'm guessing we want that to match the actual C API version.
  • There's no GPG signing, it would be easy to add but until now we haven't GPG signed composefs-rs releases so this maintains the status quo. RFE if we want to add that in a followup.

I tested all of this in my fork, seems to work, see for example https://github.com/jeckersb/composefs-rs/actions/runs/31035639379 which created jeckersb#3.

@cgwalters

Copy link
Copy Markdown
Collaborator

Honestly I am not super happy with the bootc one, having it only half-automated where it just creates the draft doesn't feel seamless enough. I think we should probably have the release driven through an issue where we edit release notes there or so.

I removed all of the app token stuff which doesn't seem to be necessary anymore? Per https://github.com/peter-evans/create-pull-request#token I think we can just use the default GITHUB_TOKEN. Not at all positive about that though.

That depends on the repo, we don't have this checked here

image

@cgwalters cgwalters left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sure we can try it though

@cgwalters
cgwalters added this pull request to the merge queue Aug 6, 2026
Merged via the queue into composefs:main with commit e8cfebc Aug 6, 2026
19 checks passed
@jeckersb

jeckersb commented Aug 6, 2026

Copy link
Copy Markdown
Collaborator Author

Also worth noting this only does the PR creation and there's a whole other workflow that does the actual release, that one (or some modified and/or rewritten equivalent) will need to be brought over here too. For the time being we can use this to generate the PR and then once it lands we can do a manual release for 0.8 so we don't delay any longer.

@jeckersb

jeckersb commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator Author

Honestly I am not super happy with the bootc one, having it only half-automated where it just creates the draft doesn't feel seamless enough. I think we should probably have the release driven through an issue where we edit release notes there or so.

I removed all of the app token stuff which doesn't seem to be necessary anymore? Per https://github.com/peter-evans/create-pull-request#token I think we can just use the default GITHUB_TOKEN. Not at all positive about that though.

That depends on the repo, we don't have this checked here
image

It does not work without the box checked, see https://github.com/composefs/composefs-rs/actions/runs/31126157009

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants